Hello,

Our product is 3V battery Operated, And fit outsides, and uses ceramic capacitor of 25V or above.

Devices are installed for 1 year.
In 1 year, in some devices, after 3-5 months, ceramic capacitors cause leakage current, So, the device's battery drains fast.
In deep sleep our device consumes 3.7uA, but when this issue occurs it consumes about 3mA-10mA.

In PCB placing, there is no mechanical strain on PCB components, all capacitors are on top side, and don't touch anything.

We found that, In 90% or above cases, capacitors values 10uF or above more causes leakage current after 4-5 months.
When measuring resistance of these capacitor, it shows about 10-100kohms.

We want to know what causes this issue, and how we can solve this?


Thank You

 

Hello,


Our product is 3V battery Operated, And fit outsides, and uses ceramic capacitor of 25V or above.


Devices are installed for 1 year.

In 1 year, in some devices, after 3-5 months, ceramic capacitors cause leakage current, So, the device's battery drains fast.

In deep sleep our device consumes 3.7uA, but when this issue occurs it consumes about 3mA-10mA.


In PCB placing, there is no mechanical strain on PCB components, all capacitors are on top side, and don't touch anything.


We found that, In 90% or above cases, capacitors values 10uF or above more causes leakage current after 4-5 months.

When measuring resistance of these capacitor, it shows about 10-100kohms.


We want to know what causes this issue, and how we can solve this?



Thank You

Almost all capacitors are susceptible to at least some level of leakage. It usually increases with temperature so for example if the circuit board resides within a hot cabinet then you should expect a higher degree of leakage. Leakage can also occur due to the things with which it is directly connected to. This includes a humid environment which can as well act as a conductive path.

So maybe start with the data sheet and move forward from there. Physically probe and inspect the circuit and obtain as many test parameters as possible. The more specific data you have about the system, the better.

 

The largest ceramic capacitor I use is 0.1uF. Your 10uF is 100 times higher.
I buy name-brand capacitors. Are yours unknown-brand from ebay?

I drew it from scratch.
Depends on the alternate and flash rates you want. You can use this nomogram from National Semiconductor to estimate values. Or you can do the calculations from the datasheet.
View attachment 318657View attachment 318658
If you make RB much much larger than RA, you'll get about a 50% duty cycle.

Yeah OK. I’m trying to simulate the front grill strobes in this video….ah guess I can’t attach a video

 

Hello,

Our product is 3V battery Operated, And fit outsides, and uses ceramic capacitor of 25V or above.

Devices are installed for 1 year.
In 1 year, in some devices, after 3-5 months, ceramic capacitors cause leakage current, So, the device's battery drains fast.
In deep sleep our device consumes 3.7uA, but when this issue occurs it consumes about 3mA-10mA.

In PCB placing, there is no mechanical strain on PCB components, all capacitors are on top side, and don't touch anything.

We found that, In 90% or above cases, capacitors values 10uF or above more causes leakage current after 4-5 months.
When measuring resistance of these capacitor, it shows about 10-100kohms.

We want to know what causes this issue, and how we can solve this?


Thank You

If your measuring a resistance across a ceramic capacitor, out of the circuit, and its reading 100 K ohms that is wrong
as an example of what to expect, look here
https://www.murata.com/en-eu/support/faqs/capacitor/ceramiccapacitor/char/0039

How are you measuring the leakage / resistance ?
not in circuit , or held in the hand I assume

Its not "just damp" by any chance ,

 

This is why some manufacturers add conformal coating to their boards.

 

If your measuring a resistance across a ceramic capacitor, out of the circuit, and its reading 100 K ohms that is wrong
as an example of what to expect, look here
https://www.murata.com/en-eu/support/faqs/capacitor/ceramiccapacitor/char/0039

How are you measuring the leakage / resistance ?
not in circuit , or held in the hand I assume

Its not "just damp" by any chance ,

We observe Battery % of on field device, if battery drain faster than expected, then we bring back device, and measure its current consumption, which is 3mA-10mA. then We remove ceramic capacitor one by one, start from battery input terminal. If current measurement down to 3.7uA by removing specific capacitor, then we measure its capacitance and resistance.
capacitance of capacitor doesn't change , but it shows resistance ranges from 10K-100K, lower the resistance, higher the current consumption, as these capacitor are power supply filter capacitor.

 

Change to a better brand, if possible change to a better dialectic.

 

We observe Battery % of on field device, if battery drain faster than expected, then we bring back device, and measure its current consumption, which is 3mA-10mA. then We remove ceramic capacitor one by one, start from battery input terminal. If current measurement down to 3.7uA by removing specific capacitor, then we measure its capacitance and resistance.
capacitance of capacitor doesn't change , but it shows resistance ranges from 10K-100K, lower the resistance, higher the current consumption, as these capacitor are power supply filter capacitor.

That's great
Sounds like you have be bunch of duff ceramic capacitors there .

 

Change to a better brand, if possible change to a better dialectic.

Please suggest best brand for ceramic capacitors

 

I leave that huge research project to you or the consultant of your choice. I don't have the time to dedicate to such a project. Perhaps others on this site have the time to dedicate to such an effort or can at least make helpful casual suggestions.

 

Please suggest best brand for ceramic capacitors

What brand do you currently use ?
what is the dielectric type ?
what is the package size
is the max working voltage of 3v you mention sufficient

 

What brand do you currently use ?
what is the dielectric type ?
what is the package size
is the max working voltage of 3v you mention sufficient

We are currently using AVX and Yageo brand mostly X7R and X5R,
We are using 0805 package and device max operated at 3.3V, but we are using 25V ceramic capacitor.

 

Please suggest best brand for ceramic capacitors

Bad ceramic capacitors - especially capacitors that turn bad at some narrow timeframe is unlikely. I think a software glitch that prevents deep sleep or other sensor glitch is happening that drains the batteries - and the capacitor is getting damaged and starts leaking when it is unsoldered from the board for testing.

 

Do you want to buy some Tantalum capacitor to try?

 

If you have some failed AVX capactors you should be able to get them to comment on the capacitor and the circuit, especially if you send them defective capacitors to evaluate. Yageo I am not familiar with.

 

Lets try some math as a sanity check. You are getting 3 ti 10 mA through a resistance of 10K to 100K. That implies at least 30V (3mA into 10K) and as high as 1000V (10mA into 100K)

If you are really using these voltages with a 25V capacitor, the problem is obvious. But I suspect there is something wrong in what you are measuring.

Edited to add: Instead you should measure the leakage current at the highest voltage you expect on the suspected capacitor.

Bob

 

The largest ceramic capacitor I use is 0.1uF. Your 10uF is 100 times higher.
I buy name-brand capacitors. Are yours unknown-brand from ebay?

I have used MLCCs of up to 100 uF. Get with the times!

Bob

 

A company I used to work for had a similar issue with ceramic caps developing high leakage current after a period of time (it was a battery powered alarm sensor) and it turned out to be the caps getting damaged when the boards where being removed from the platter.

We ended up having to work with the cap manufacture who examined the faulty caps for use and found the issue

Interestingly we did find that if we heated the caps or removed them from the PCB they would measure ok until they cooled down

 

For those of us not familiar with the term, what is "the platter"?

 

For those of us not familiar with the term, what is "the platter"?

A platter is the sheet of pcb,s that came from the PCB assembler witch had been scored but not separated into single units.
it was use separating then that was causing the damage to the caps.